Pricing efficiency, trading behavior, spillover effects and market integration: a study of emerging markets ETFs

Abstract

Given the increased interest of the international investment community in investing in stock markets of emerging economies over the recent years, this thesis aims at providing reliable empirical answers to key questions raised by Exchange Traded Funds (ETFs) investors. ETF investments are regularly sought to provide diversification benefits and possibly enhanced returns in their portfolios. One of the questions regards the pricing efficiency of these markets and the potential for investors gaining any substantial abnormal returns by exploiting any possible pricing inefficiencies. Another question relates to the trading behavior of ETFs covering emerging stock markets. Going further, ETF investors in emerging markets need to know whether there are any really meaningful diversification benefits from investing in emerging markets and how these markets interrelate to developed markets. Finally, investors who choose less traditional ETF products, such as the leveraged and inverse leveraged ETFs, to obtain access to emerging stock markets need to know the merits and pitfalls of these products as well as their comovement with developed markets. The thesis examines these significant researching issues by focusing on ETFs which trade on the stock market of the United States but are invested in indices from emerging country or region markets. Six chapters are contained in this thesis, each of which tries to contribute to answering the researching questions raised in this study. Moreover, chapters two, three and four use a common sample of forty iShares which are listed in the New York Stock Exchange but track country or regional indices from Americas, Europe, Asia and South Africa. Chapter five employs a sample of 30 ETF pairs (i.e., 30 US-listed ETFs and 30 ETFs listed in the Hong Kong Stock Exchange and the Singapore Stock Exchange), each pair of which tracks the same stock index. Finally, chapter six uses a sample of twelve leveraged and eleven inverse leveraged ETFs covering country or regional emerging market indices. Chapter one constitutes an introduction to ETFs in order for the reader to become familiar with this relatively unknown to the general public investment tool. A brief historical analysis of the ETF market is provided. Then, the characteristics and benefits of ETFs are discussed along with the several types of ETF structures and the various trading strategies that can be implemented with ETFs. Next, the chapter describes the empirical findings of the literature relating to the competition between ETFs and traditional mutual funds, the tracking ability of ETFs and the factors that can affect their replication efficiency, and the relationship of returns with the divergence between the trading prices and net asset values (NAVs) of ETFs.Chapter two studies the weak-form efficiency of emerging markets ETFs in order to answer whether investors could execute any profitable strategies based on the past returns of ETFs. Several parametric and non-parametric empirical tests are applied in this respect. These tests first include estimating the autocorrelation and serial correlation in ETF return time series. Next, runs tests are applied to evaluate the randomness in ETF return series. Finally, efficiency is investigated by applying three alternative types of variance ratio tests to assess whether ETF prices follow a random walk. These three tests are the Lo and MacKinlay (1988) variance ratio test, the ranks-based and signs-based variance ratio of Wright (2000) and the Chow and Denning (1993) variance ratio test. Overall, the results of the tests reveal that the pricing of emerging ETFs listed in the US stock market is efficient at the weak level. This means that the already publicly available information incorporated in the prices of ETFs does not provide investors with significant chances of gaining material abnormal returns.Chapter three aims at informing investors about the trends in the trading behavior of emerging markets ETFs so that they can make informed investment decisions. The chapter focuses on the return and volatility of ETFs during trading and non-trading hours, their performance relative to market performance, the tracking error and persistence in tracking error, the divergences between the trading prices and net asset values of ETFs, the persistence in these inefficiencies and the reaction implications for ETF pricing due to the existence of these inefficiencies. The results show that the ETFs slightly underperform their benchmarks being also more volatile than them. Moreover, ETFs do not produce any excess return relative to their benchmarks. In addition, it is shown that the trading prices of ETFs are significantly affected by the trends in the US stock market and the Fama and French market capitalization and book-to-price factors but not their NAVs. Going further, the results show that the tracking error of these funds is substantial and that the tracking error tends to revert from day to day. Furthermore, the analysis shows that the average ETF trades at a persistent premium to its NAV. Finally, the analysis reveals that the trade price return of ETFs is positively and negatively related to NAV returns and one-day lagged premium, respectively.Chapter four wants to inform investors about the interrelationship between developed and emerging markets by investigating the return and volatility spillovers between the US ETF market and the underlying emerging stock markets using advanced econometric and correlation analysis techniques. At first, a comprehensive correlation analysis containing the Pearson’s correlation coefficient and the conditional constant and dynamic conditional correlation coefficients is applied to assess comovement between the two markets. Then, a VARMA model, a VARMA-GARCH model and a VARMA-EGARCH model are performed to evaluate the spillover effects on returns. Finally, volatility spillovers are examined with an Augmented GARCH model, an ARMA-GARCH model, an ARMA-EGARCH model, a scalar-BEKK model, and an ARMA-scalar-BEKK model. The empirical findings reveal a high degree of comovement between the US ETF market and the underlying emerging stock markets while significant bilateral return and volatility spillover effects between ETFs and benchmarks are accentuated too. Chapter five assesses the price integration between the US and emerging Asian ETF markets from a “Law of One Price” perspective. In doing so, the chapter applies a wide range of integration tests, such as autoregressive moving average models with autoregressive conditional heteroskedasticity innovations, dynamic conditional correlation analysis, Granger causality tests and autoregressive distributed lag cointegration tests. The main purpose of this chapter is to answer whether there are diversification or other benefits to investors seeking exposure to emerging markets by simply switching from US-listed to similar Asia-listed ETF products and vice-versa. The findings accentuate a high degree of integration between the US and the emerging ETF markets examined. In other words, the empirical results show that the Law of One Price is not rejected and, consequently, there are no beneficial diversification strategies, at least in performance terms, that can be implemented by investors by investing in ETFs written on the same benchmarks but listed in different stock exchanges. Chapter six seeks to shed light on the pros and cons of less traditional ETF products available to investors who seek exposure to emerging markets and their comovement with the developed markets. The focus of the chapter is on the performance and volatility of leveraged and inverse leveraged ETFs tracking indices from emerging markets. In particular, the short-term and long-term performance vis-à-vis the return of targets is investigated along with their volatility, the volatility persistence, the spillover effects on returns and the transmission of volatility between ETFs and indices. The results demonstrate that the average leveraged ETF can deliver its performance target over a weekly period at a maximum whereas the respective inverse leveraged ETF can do so only over a two-day period. When it comes to risk, the findings reveal that the volatility of leveraged ETFs is quite aligned to the volatility of targets also being significantly persistent through time. Moreover, it is found that the leverage effect, which says that the volatility of a stock increases when equity prices fall, applies to leveraged ETFs but the opposite trend is the case for inverse ETFs. Finally, significant bi-directional spillover effects on return and volatility between ETFs and benchmarks are revealed. This thesis contributes to the literature in several ways. In particular, to the best of our knowledge, the study conducted in chapter two on the weak-form efficiency of ETF markets is the first one to focus on emerging markets ETFs. The results of the study are strongly supportive of the validity of the efficient market hypothesis and contribute to the long-lasting debate of whether capital markets are efficient and whether past returns can be indicative of future returns with new insights from a market which has attracted a might interest from the investment community but has not been investigated yet.The study on the trading behavior of emerging markets ETFs in chapter three plays a supplementary role to the examination of market efficiency in chapter two. Investors in emerging markets ETFs need to know whether they stand any chances of building profitable strategies based on the information incorporated in the historical prices of ETFs but they also need to be aware of the key elements relating to and affecting their investments in these products. By investigating ETFs’ return, risk, and tracking and pricing efficiency, this study provides investors with all the significant information they need to have so as to make informed investment decisions.Going further, the study on spillover effects on return and volatility of emerging markets ETFs in chapter four is just the second study in the literature which examines these issues focusing on developed and emerging markets and use data from the ETF industry. The first study on the interrelationship between emerging and developed ETF markets was that of Chen and Huang (2010). However, to our view, the study conducted in chapter four stands as a major expansion of the work of Chen and Huang (2010) given that we use a sample of forty pairs of ETFs and underlying indices while the previous study used only nine. The significantly bigger sample used entails that a wider section of emerging markets is covered compared to the study of Chen and Huang (2010) and, consequently, our results are more representative of emerging markets as a whole. Another contribution is that we use a significantly wider range of econometric and statistical techniques than Chen and Huang (2010), who only employed the VARMA-GARCH and VARMA-EGARCH models. The wider range of methodological techniques ensures that our results are absolutely solid without being subject to the assumptions of the models used to produce them. One significant contribution of the study on price integration between developed and emerging ETF markets in chapter five relates to the kind of the sample used in the study. In particular, this is the first study focused on emerging markets which uses pairs of ETFs written on the same indices but listed in one developed and one emerging stock exchange to assess market integration. On the other hand, the common approach in the literature relates to the usage of major stock indices from several developed and emerging markets or commodities traded in developed and emerging markets. Moreover, it can be said that the literature concerning the integration of developed and emerging ETF markets are relatively scarce and, consequently, our study is a major contribution to the relevant literature given the significant empirical results obtained. Finally, the study on emerging markets leveraged ETFs in chapter six is the first study in the relevant literature on leveraged ETFs to focus exclusively on emerging markets. In addition, it is just the second study in the literature on leveraged and inverse leveraged ETFs which investigates the spillover effects on return and volatility between these ETFs and their benchmarks.

Δεδομένου του αυξημένου ενδιαφέροντος της διεθνούς επενδυτικής κοινότητας για τις αναδυόμενες οικονομίες κατά τα τελευταία χρόνια, η παρούσα διατριβή έχει ως στόχο να παράσχει αξιόπιστες εμπειρικές απαντήσεις σε βασικά ερωτήματα που τίθενται από επενδυτές οι οποίοι χρησιμοποιούν τα λεγόμενα Διαπραγματεύσιμα Αμοιβαία Κεφάλαια (Exchange Traded Funds-ETFs) για να αποκτήσουν πρόσβαση στις αγορές αυτές. Ένα από αυτά τα βασικά ερωτήματα αφορά στην αποτελεσματικότητα της αγοράς και στη δυνατότητα που ενδεχομένως υπάρχει για τους επενδυτές σε ETFs αναδυομένων αγορών να αποκομίσουν σημαντικές υπερβάλλουσες αποδόσεις από την αξιοποίηση οιασδήποτε αναποτελεσματικότητας στην τιμολόγηση αυτών των προϊόντων. Η εμπορική συμπεριφορά των ETFs ενδιαφέρει επίσης τους επενδυτές. Η δυνατότητα εφαρμογής επιτυχημένων στρατηγικών διαφοροποίησης ρίσκου είναι ένα άλλο ζήτημα ιδιαιτέρου ενδιαφέροντος για τους επενδυτές καθώς και ο βαθμός συσχέτισης των αναδυομένων με τις αναπτυγμένες κεφαλαιαγορές. Τέλος, οι επενδυτές που επιλέγουν ETFs νέου τύπου, όπως τα λεγόμενα ETFs μόχλευσης, προκειμένου να αποκτήσουν πρόσβαση στις αναδυόμενες αγορές, χρειάζονται πληροφόρηση σχετικώς με τα πλεονεκτήματα και τα μειονεκτήματα αυτών των προϊόντων καθώς και για τη συσχέτιση τους με τις αναπτυγμένες αγορές. Τα αποτελέσματα της παρούσης διατριβής προσφέρουν αξιόπιστες απαντήσεις στα παραπάνω βασικά ερωτήματα. Συγκεκριμένα, αποδεικνύεται ότι η υπό εξέταση αγορά ETFs είναι κατά βάση αποτελεσματική και άρα οι επενδυτές δεν έχουν δυνατότητα για σημαντικές αποδόσεις πέραν των μέσω αποδόσεων της αγοράς. Επιπλέον, η απόδοση των ETFs είναι ελαφρώς χαμηλότερη από αυτή των υποκείμενων δεικτών ενώ, πέραν από τον υποκείμενο δείκτη, μπορεί να επηρεάζεται και από παράγοντες της αμερικανικής κεφαλαιαγοράς όπως για παράδειγμα η απόδοση του δείκτη Standard and Poor’s 500. Περαιτέρω, αποδεικνύεται ότι η χρηματιστηριακή τιμή του μέσου ETF διαφέρει σημαντικά από την καθαρή εσωτερική του αξία. Ένας υψηλός βαθμός συσχέτισης μεταξύ των αναπτυγμένων και αναδυομένων αγορών ETFs αναδεικνύεται επίσης καθώς και σημαντικές επιδράσεις στην απόδοση και τον κίνδυνο από τα ETFs στους υποκείμενους δείκτες και αντιστρόφως. Τέλος, αναδεικνύεται ότι τα ETFs μόχλευσης μπορούν να επιτυγχάνουν τους επενδυτικούς τους στόχους. Η βασική συνεισφορά της διατριβής αυτής στη σχετική βιβλιογραφία αφορά στο γεγονός του ότι εξετάζεται μια αγορά υψηλού ενδιαφέροντος, όπως είναι αυτή των ETFs αναδυομένων αγορών, η οποία είναι υπό-μελετημένη. Η παροχή αξιόπιστων εμπειρικών αποτελεσμάτων είναι μια άλλη συνεισφορά της παρούσης μελέτης. Από μεθοδολογικής απόψεως, οι οικονομετρικές τεχνικές που εφαρμόστηκαν είναι αρκετά σύγχρονες και συνάδουν με την υπάρχουσα σχετική βιβλιογραφία. Τέλος, τα δείγματα που εξετάστηκαν είναι αρκετά αντιπροσωπευτικά της συνολικής αγοράς, καθώς καλύπτουν ένα σημαντικό εύρος των διαφόρων αναπτυσσόμενων οικονομιών, αλλά και η επιλογή τους είναι αρκετά πρωτοποριακή συγκριτικά με τις συνήθεις πρακτικές επιλογής δειγμάτων στη σχετική βιβλιογραφία.